# Switching Solar Panels in Series and Parallel

1. ## Switching Solar Panels in Series and Parallel

Hi all,

I want to build a series/parallel combination of solar panels.

The panels I propose to use have to be connected in series (6S configuration) to attain a voltage of around 200V DC to drive a motor.
While I also have a battery bank to be charged at 48V DC. I want to connect the panels in a 2S3P configuration.

If I use a MOSFET to do the switching, the MOSFET pins are connected to the same potential hence the MOSFET doesnt switch ON. I am not sure if I can use an SSD to do the switching.

Can anybody help or guide me on the above problem on if I can use Mosfets the way I have described them?
Thank you

•

2. ## Re: Switching Solar Panels in Series and Parallel

Using MOSFETs is quite difficult because the gates have to be lifted close to drain voltage of each device. As the source and drain connections are all at different voltages, you need quite a complex gate drive circuit to provide a different voltage for each device.

If you can live with a simple dual-pole change-over relay there is a much simpler solution. I can't draw diagrams at the moment so forgive the description:
1. from 0V (the negative line) connect panels 1 & 2 in series.
2. from the top (positive end) make two connections, one to the anode of a diode, the other to the NC pin of one relay pole. Diode cathode goes to the output point.
3. the NO pin of the same relay pole goes to 0V, the common goes to the negative side of panels 3 & 4 in series.
4. repeat the above using the other relay pole for panels 5 & 6.

You should have three diodes in total, each with cathodes meeting at the output point.

With the relays un-energized (common and NC connected), the three pairs of panels are in series with the output taken from the top through diode 3.
With the relay energized (common and NO connected), the three panels negative side are joined to 0V and the output is taken in parallel from diodes 1, 2 and 3.

Technically you can dispense with the third diode but it might cause problems with unequal voltages from the panels. Also, if you really want a MOSFET solution, you can replace the relay contacts with four MOSFETs in that configuration and it reduces the gate signals needed to three because two of them switch simultaneously to 0V when simulating the NO relay contacts. You might be able to provide the 'high' gate voltages by wiring a resistor between gates and drains and adding an opto-coupler to pull the gates down to respective source voltages.

Let me know if you understand, I may be able to draw a schematic when I'm back at my desk tomorrow.

Brian.

•

3. ## Re: Switching Solar Panels in Series and Parallel

Hi Brian,

Thank you for your solution. The problem with using diodes and relays is that the panels are rated at 315W. So the current at Vmp is around 8A. When working in 2S3P configuration, the diodes and relays used for switching will have to handle around 24A of current at 70V. I am not sure if I can find an economical diode or relay capable of handling that amount of current at that voltage. Please suggest one if you know.

I was thinking of using MOSFETs because that way the solution can be economical. IRF540N seems to do the job with Vds maximum at 100 V and Id at 33A.

Right now I am proposing to use a 2 pole changeover with one end connected to the motor while other connected to the inverter as shown in the schematic.

When I operate switch 1 and 2(in the switchover they are connected, hence operating switch 1 operates switch 2 also and disconnect switches 3 and 4), the 6 panels are in series which give the required 200 V to drive the motor. Simultaneously, the inverter is connected between panels 1 and 2, so it also keeps on charging the battery in 2S configuration.

When I operate Switch 3 and 4(the switchover disconnects switches 1 and 2), the panels are connected in 2S2P configuration and end up charging the batteries through the inverter at 2x current. The downside is that with this, I can utilise only 4 panels for charging the batteries. I guess a 3 pole changeover can eliminate this problem too.

Would request your take on this.

When

Originally Posted by betwixt
Using MOSFETs is quite difficult because the gates have to be lifted close to drain voltage of each device. As the source and drain connections are all at different voltages, you need quite a complex gate drive circuit to provide a different voltage for each device.

If you can live with a simple dual-pole change-over relay there is a much simpler solution. I can't draw diagrams at the moment so forgive the description:
1. from 0V (the negative line) connect panels 1 & 2 in series.
2. from the top (positive end) make two connections, one to the anode of a diode, the other to the NC pin of one relay pole. Diode cathode goes to the output point.
3. the NO pin of the same relay pole goes to 0V, the common goes to the negative side of panels 3 & 4 in series.
4. repeat the above using the other relay pole for panels 5 & 6.

You should have three diodes in total, each with cathodes meeting at the output point.

With the relays un-energized (common and NC connected), the three pairs of panels are in series with the output taken from the top through diode 3.
With the relay energized (common and NO connected), the three panels negative side are joined to 0V and the output is taken in parallel from diodes 1, 2 and 3.

Technically you can dispense with the third diode but it might cause problems with unequal voltages from the panels. Also, if you really want a MOSFET solution, you can replace the relay contacts with four MOSFETs in that configuration and it reduces the gate signals needed to three because two of them switch simultaneously to 0V when simulating the NO relay contacts. You might be able to provide the 'high' gate voltages by wiring a resistor between gates and drains and adding an opto-coupler to pull the gates down to respective source voltages.

Let me know if you understand, I may be able to draw a schematic when I'm back at my desk tomorrow.

Brian.

•

4. ## Re: Switching Solar Panels in Series and Parallel

8A Schottky diodes are plentiful and cheap. The most current any diode has to carry is 8A, when in parallel mode, each pair of panels has it's own diode so none of them carry the combined current of your 2S3P configuration. Sorry about the hand drawn sketch!

Brian.

5. ## Re: Switching Solar Panels in Series and Parallel

Good luck finding an "inexpensive" relay with contacts that are rate at 250 Volts DC at 10 Amps.
A relay with contacts that are "AC Rated" only, will not work.

Why not just leave the panels always connected as "6S" and
use a "250 Volt Input to 48 Volt @ 30 Amp output" MPPT Charge Controller for the batteries?

Connecting the Solar Panels directly to the battery bank without a Charge Controller is a bad idea.

OR connect as 3 sets of "2S"
and Run the motor from the battery bank using a 48v-to-200v Boost Converter?

--[[ ]]--